Who knows what the impact of a new feature will be. Will people understand it? Are we going to get swamped with support requests? Will it be a drag on performance? You can always count on uncertainties to scare you off. There are plenty of “What ifs” that can kill a good idea through fear, doubt, and uncertainty.
What you need to climb that wall is a willingness from someone to accept the risks and pay the price. The champion. He who wants this bad enough to be willing to deal with the initially confused customers, answer support emails related to the feature, and put in the work to make it fast and solid.
With a champion in place, most of the counter-arguments lose their appeal as they’re usually born from a perceived discrepancy between cost and reward. As in “sure you want this, but I’m going to be the one who has to deal with the mess”. When cost is assigned to the bulls, the perceived discrepancy of the bears dissipates.
The working title for this story: “How 37signals adopted OpenID”